Nutritional Requirements During Training for Special Operation Forces – An Observational Study

Background: Responses to exercise training can vary greatly between individuals. For special operation forces, low responses to training can hamper performance. In this study, we objectively measured strength and tness during special operation forces training, and assessed potential determinants of the training response. Methods: Twenty subjects were enrolled, and measurements were taken before and after a 9-week training program. Muscular strength was measured as one-repetition-maximum on four instruments, and physical tness by the Cooper-test. Body composition was measured using deuterium dilution, physical activity by accelerometry and diet quality by food records. Level of signicance was p<0.05. Results: During the 9-week training period, body strength increased by 0.33±0.24 N/kg (+7%, P<0.001, and physical tness increased by 3.5±3.4 mL/min/kg (+6%, P=0.001). Gains in strength were inversely associated with strength at baseline, and positively with activity intensity during the training program. We observed no effect of training on body weight, but body composition was signicantly different at follow-up as compared to baseline (16.9±2.5% to 14.9±2.5% body fat, P=0.03). Energy intake was 4491±506 kcal/d and energy balance was -243±306 kcal/d (P=0.04). Average physical activity level was 2.6±0.2 and the average duration of moderate-to-vigorous physical activity was 5:53±0:36h. Over time, physical activity did not change signicantly. After adjustment for underreporting, intakes of vitamin C and D were insucient on average and for most participants. Conclusions: Improvements in strength were modulated by strength prior to the intervention, and moderate-to-vigorous physical activity during the training. Thus, compensatory declines in physical activity may hamper the effectiveness of the exercise program. composition-data. P-values the statistical signicance for the


Introduction
Military performance is crucially dependent on soldiers physical tness and health. Low mass and strength of the musculoskeletal apparatus increases the risk for injuries and fatalities [1,2]. Impaired metabolic health may negatively affect endurance, resiliency, and recovery [3,4].
To increase physical tness and strength, all soldiers undergo baseline training and maintain a schedule of regular exercise thereafter. Between military personnel, physical tness requirements differ by position and requirements for special operation forces are higher as compared to other task forces. To achieve increased physical tness, these soldiers follow another, specialized training in which training load is higher, both in duration and intensity.
Responses to exercise training can vary greatly between individuals [5]. Factors that may explain such variability may include training intensity [6], nutritional adequacy and quality [7], and compensatory changes in habitual physical activity and nutrition [6,8]. In a relatively standardized setting of location, training schedules, and diets, it is important to determine whether such variability in exercise reponse is also observed in military recruits and to identify factors that modify training effects.
To gain insight into the effect of special forces operations training and predictors of this response, we performed a study in which we assessed physical tness and strength, as well as body composition, diet quality, and physical activity in soldiers following an 9-week training.

Aims, Design and Setting
Before and after the 9-week training program, measurements of strength, tness, and body composition were performed. During the training period, physical activity and diet quality were assessed. During the base training, recruits spend ve days a week at the military basis, performed their training and receive standard military diets. During weekends, recruits were allowed to travel, e.g. home, with no instructions for physical activity and diet.

Participants
Participation in this study was offered to recruits who were to follow the base training of the Special Forces of the Royal Dutch Army. Twenty, healthy subjects volunteered to participate in this study between March and May 2016.
10 participants completed measurements of body composition before and after training: among the 10 dropouts, Four subjects did not complete the study, two subjects left the military training program, one subject quit the study because of the experienced high load of the study and one did not perform tness testing after the study, 5 measurements resulted in unreliable body water-estimates, and one participant did not collect the post-dose urine sample before training. Anthropometrics, strength and tness were not different between those who completed measurements and those who did not (Table 1). P-values refer to comparison of baselines values between participants with complete data and incomplete data at follow-up (for cohort for primary outcomes: n = 16 vs n = 4, for cohort for secondary outcomes: n = 10 vs n = 10, data of drop-outs not shown). Strength and tness are expressed per kg body weight.

Muscular strength
Muscular strength was measured as one-repetition-maximum on the chest press, leg press, vertical traction, and shoulder press. Strength was calculated as the sum of forces on four exercises divided by body weight.

Physical tness
Physical tness was assessed by the Cooper-test [9]. The Cooper-test assesses the distance participants are able to run in 12 minutes. Physical tness is estimated as age-and sex-speci c function of the achieved distance.

Body composition
Anthropometrics of the subjects were obtained at the second day and at the last day of their base training.
Body weight was measured with minimal clothing (e.g. underwear) after an overnight fast. Body composition was measured using Deuterium dilution according to the Maastricht Protocol [10]. Before the subjects went to bed, they collected a baseline urine sample. Immediately thereafter, subjects ingested 70 mL of the deuterium solution. The following day, approximately 8 hours after consuming the isotope dilution, a sample of their second morning urine was collected. In-between ingestion of the deuterium solution and collecting the second morning urine samples, the subjects were not allowed to consume anything. Total body water was calculated by the Plateau-Method, and fat-free mass was calculated as total body water divided by 0.73, assuming 73% hydration of fat-free mass. Fat mass was calculated as difference between body weight and fat-free mass.

Energy homeostasis
Energy intake was calculated from self-report by dietary records ('Reported'), and using the energy intakebalance method ('Calculated'). The intake-balance method utilizes the rst law of thermodynamics, and calculates energy intake as the sum of energy expenditure and changes in body energy stores. Total energy expenditure was estimated using accelerometry. Changes in body energy stores were calculated as the difference between fat mass and fat free mass in the rst and last week of the study, multiplied by their respective energy densities of 9300 kcal/kg fat mass and 1100 kcal/kg fat-free mass [11].

Dietary quality
A daily food record was used to assess dietary intake. Food records were completed by the participants every weekday and during one weekend by the participants.
During week days, food and drinks were supplied to the participants, allowing speci c knowledge about their diet composition. During weekends, diets were ad libitum. Dietary intake was quanti ed as energy content, macronutrient composition (as percentage of total energy), and adequacy of micronutrient and vitamin intake (as % of the Recommended Daily Allowance) using the NEVO-table.

Physical activity
Physical activity was monitored using waist-worn accelerometers (ActiGraph GT3X, Actigraph, Pensacola, FL, USA). Physical activity was recorded during the entire training period of nine weeks. To minimize burden, every participant wore the accelerometer for 4 out of 9 weeks; one week in each period: week 1-2, week 3-4, week 5-6-7, and week 8-9. For each period, accelerometers were randomly assigned to 10 participants; the remaining 10 wore the accelerometer during the following week. Data of non-wear weeks was linearly imputed from the week before and after. Accelerometers were worn 24/7 and only detached prior to water activities, e.g. swimming, showering. The subjects wore the accelerometers only for one weekend to monitor physical activity behaviour at home.
The ActiGraph is a compact (3.8 × 3.7 × 1.8 cm) and lightweight (27 g) device that has a rechargeable lithium polymer battery. It measures acceleration in three planes: mediolateral (X), vertical (Y), and anteroposteral (Z). The sampling rate of the accelerometer was set at 30 Hz. A low frequency extension lter was used to analyze the data. The acceleration data of the three planes from each minute was used.
The vector magnitude was calculated as followed: √(Axis 1 2 + Axis 2 2 + Axis 3 2 ). The Freedson Vector Magnitude 3 cut-off values were used to calculate the intensity of the physical activities per minute [12]. These cut-off values were based upon the counts per minute (CPM

Statistical analysis
Results of all the parameters were expressed as mean ± SD. Normality was con rmed for all variables using Shapiro-Wilk-tests. Changes over time, e.g. body weight and composition, were analyzed using paired samples t-tests, baseline differences between subjects for secondary outcome analysis and excluded participants due to incomplete follow-up data were compared using unpaired t-tests.

Body weight and composition
During the 9-week training, body weight did not change ( Fig. 2A). In 10 participants with body composition data at follow-up, we observed no change in body weight either, but body composition was signi cantly different at follow-up as compared to baseline (16.9 ± 2.5% to 14.9 ± 2.5% body fat, P = 0.03, Fig. 1B). Fat mass decreased signi cantly (-1.8 ± 2.2 kg, P = 0.03), whereas fat-free mass increased, yet not statistically signi cant (+ 1.2 ± 2.0 kg, P = 0.09).
Protein intake was 2.35 ± 0.25 g/kg body weight/day.
Based on self-reported dietary intake, intakes of Vitamin B12, C, D, and salt were signi cantly lower than the Recommended Daily Allowances (Table 2). After adjustment (intake divided by % underreporting), only intakes of vitamin C and D were insu cient on average and for most participants.

Determinants of improvements in strength and tness
The increase in strength during the base training was negatively associated with strength (r=-0.62, p = 0.01), and tness (r=-0.52, p = 0.04) at baseline. Furthermore, the increase in strength during 9 weeks of training was associated with an increase in moderate (r = 0.56, p = 0.03), vigorous (r = 0.59, p = 0.02), and very vigorous-intensity activity (r = 0.66, p = 0.01) during training. Lastly, larger gains in strength (adjusted for body weight) associated with larger energy de cits (r=-0.66, p = 0.04, n = 10), and losses in fat mass (r=-0.66, p = 0.04, n = 10), but not with changes in fat-free mass (N.S.). If strength was adjusted for fat-free mass instead of body weight, these associations disappeared. Changes in tness were not associated with any assessed variable.

Discussion
In the present study, we report the improvements in tness and strength observed during a 9-week special operation forces training in soldiers of the Dutch Army. This is the rst study to objectively measure physical activity and activity intensity using accelerometry during the training period. Thereby, this study allows to assess whether improvements in tness and strength are affected by the intensity of physical activity during the training period. In addition, we measured body composition and collected dietary records to determine nutritional adequacy during this high-intensity training program.
The present study demonstrates that the 9-week special operation forces training achieved the expected improvements in physical tness and whole-body strength (+ 6-7%). Adjusting for individual changes in fat-free mass, largely composed of muscle mass, the increase in tness and strength were small (+ 3%), but remained signi cant. Thus, strength and tness improved likely as a result of increased muscle mass and quality and respiratory capacity. The inter-individual range in response to the training program was 25%. For strength, the increase negatively associated with strength prior to training, demonstrating that strong individuals bene t less from the training program. For tness, no such association was observed and thus, the prescribed training achieves improvements in tness even in t individuals.
The training regimen in this observational study imposed an average physical activity level of 2.6, which is high. Inter-individual differences in physical activity during the training program, in moderate, vigorous, and very-vigorous intensities, associated positively with changes in strength. Thus, this study suggests a dose-response relationship with intense activities, but not with low-intensity activities. The training-effect on tness was not determined by the variability in physical activity levels during the 9-week training. Of note, maintaining a physical activity level at this level for > 60 days approximates to 67% of the maximal human physical capacity, which has been estimated using studies of long-term endurance capacity, e.g. artic trekking, ultra-marathons and mountaineering [15]. Importantly, these estimates are derived only from week-days, thus training days, and physical activity during the weekends was signi cantly less (-8 hours).
At the observed physical activity level, we estimated energy intake requirements during the training to be 4773 kcal/d, or ~ 60 kcal/kg body weight per day. These estimates con rm energy requirement studies using gold-standard methodology, that is doubly labeled water [7,16,17]. Importantly, accelerometry measures body movements and energy requirements are estimated as a function of body movements and body weight. During many activities, soldiers were required to carry additional weights, e.g. 20 kg-backpacks, and thus physical activity levels, energy requirements and energy intakes were likely higher than reported.
While the imposed training program did not affect body weight, soldiers lost 1.8 kg of fat mass, while gaining 1.2 kg of fat-free mass. The gains in fat-free mass are comparable to other exercise training studies [18,19]. Based on the changes in body composition, we estimated that soldiers consumed 330 kcal/d less than they expended. Energy balance was calculated based on changes in body composition and is therefore not affected by confounding by carrying additional weights. Despite being in a small energy de cit, dietary records demonstrate su cient intake of proteins, the primary dietary driver of changes in muscle mass and strength during exercise training. In contrast to other studies [20], gains in muscle mass during an energy de cit were not associated with protein intake, which is likely because protein intake was su cient in all individuals. The association between changes in strength and energy de cit is explained by changes in fat mass, because it is a factor in calculating energy balance and a denominator in calculating strength. If strength is adjusted for fat-free mass only, changes in strength do not associate with energy balance.
We did not observe any associations between changes in tness or strength with diet quality, i.e.
adequacy of nutrient intake. Nevertheless, after adjustment for underreporting of energy intake (~ 30%), we observed signi cant de ciencies in vitamin C and D intake. Although these de ciencies did not cause any adverse effects on outcomes of this study, others have shown detrimental effects of vitamin de ciencies on physical function, metabolic health and bone quality [21][22][23].
This study is the rst to use objective measurements of physical activity during high-intensity exercise training for special forces operations. In light of the intense training, burden to the participants of this study had to be minimized, and therefore, collection of this data, and in addition of diet and body composition data, is exceptional. Due to the specialized nature and target population, it is limited in its subject size, which is small, and the burden of the study in addition to the demanding training may explain the missing data. However, the results of this study, despite preliminary in nature, are valuable to inform optimization of training strategies.

Conclusions
The present study demonstrates e cacy of training programs in special operation forces to increase physical strength and tness. Improvements in strength, but not tness, were modulated by strength prior to the intervention, and moderate to vigorous physical activity during the training. Compensatory declines in physical activity may hamper the effectiveness of the exercise program. Vitamin C and D intake requirements were not met and thus might require supplementation. For future studies, the use of doubly labeled water in combination with accelerometry, as well as careful individual adherence to the training regimen is advised to explore variable responses in more detail. Declarations